© 2002 Murray Turoff and Harold A. Linstone
VII.C. Group Communication through Electronic Media *
Fundamental Choices and Social Effects
ROBERT JOHANSEN, RICHARD H. MILLER, and JACQUES VALLEE
How will a given medium of communication affect the way in which groups of people
communicate? What are the most promising near future directions for research
considering this question?
Our own incentive for exploring these issues began with a more specific concern
about the probable social effects (and utility) of communication through a
computerized conferencing system called FORUM, which is now under development at
the Institute for the Future. The starting point for our inquiry was to consider
computerized conferencing as a medium of communication, just as the telephone and
face-to-face conversations may be considered media of communication. Not
surprisingly, the criteria for evaluation of a medium of communication typically
involve (either consciously or unconsciously) comparison with other media. Since the
medium most familiar to the majority of us is face-to-face communication, there is an
inherent tendency -for this to become the standard of judgment.
One needs to exhibit great care when doing this, since computerized conferencing
and other telecommunications media are not necessarily surrogates for face-to-face
communications. It seems more likely that each medium will have its own inherent
characteristics which should not be expected to mimic face-to-face patterns. On the
other hand, comparison with -face-to-face communication is often crucial in order to
understand a new medium: While most of the work in this area to date has been applied
to conferencing media s u c h as TV and voice systems, some of it has direct bearing on
any future work in the computerized conferencing area. For instance, Anna CaseyStahmer and Dean Havron developed a mathematical ratio to aid in assessing
teleconferencing systems [1]. In this study, each system under assessment involved
groups of people gathered at stations and communicating with groups at other stations.
An analysis was made of the amount of electronically mediated communication between
stations and the communication within the face-to-face groups at each site. The point was
to look at the ratio of between-station communication over within-station
communication. This ratio has offered interesting data in this case, but needs to be used
*
Reprinted by arrangement with original publisher from Educational Technology
Magazine, August, 1974, Vol. 14, No. 8. Copyright © 1974 Educational Technology
Publications, Englewood Cliffs, New Jersey.
This is a paper from the FORUM project at the Institute for the Future. FORUM i s
being developed and evaluated by a research team composed of Roy Amara, Hubert
Lipinski, Ann McCown, Vicki Wilmeth, Thad Wilson, and the authors of this paper.
This research is supported by the National Science Foundation under Grant GJ-35
326X.
508
Computer/Future of Delphi: Group Communication
509
with great care in order to avoid the assumption that face-to-face communication is the
ideal medium.
In turning to the literature of group communication, however, one does not readily
discover general principles or procedures which are easily adopted as "standard." Instead
we find a very scattered literature-often parochial and littered with jargon-which is
impressive in its lack of coordination. Individual researchers (and often "schools" of
thought) provide rich and provocative information within strikingly narrow frames of
reference. Also, the social dynamics which have been explored in these research efforts
are concentrated almost exclusively on face-to-face communication. As evidence, one
finds only six entries dealing with media other than face-to-face among the 2,699-entry,
generally acclaimed bibliography on small-group research by McGrath and Altman [2].
Beyond the literature of face-to-face group process research, very little has been done
which attempts to apply derived principles of face-to-face group communication to other
media.
In 1963 Alex Bavelas offered this summary appraisal of the research in face-to-face
communication as it relates to research in electronically mediated group communication:
In consequence, the findings are, in most cases, only remotely related to teleconferencing. The significant contribution of this work lies instead in the methods and
techniques of quantitative study that have been developed, and in general hypotheses
about social process in terms of which specific propositions relating to teleconferencing
may be formulated [3].
Bavelas went on to say: "It appears that published information bearing directly on
teleconferencing is practically nonexistent" [4]. Thus it is clear that most of the directly
relevant research has been done within the last ten years-with the added comment that
Bavelas' observation has not changed radically since the time that he made it.
Certainly the literature of group process is broad and provocative, and the potential
for transfers into communication research seems real-though obviously complicated by
multiple factors. Alex Reid, while recognizing this fact, offers an optimistic view of nearfuture possibilities: "There seems every opportunity for a fruitful transfer of both theory
and experimental method from social psychology to telecommu nications engineering, a
transfer that will be particularly valuable as the telecommunications system moves away
from simple one-to-one voice communication. toward more sophisticated visual and
multi-person systems" [5].
One of the first research efforts which considered teleconferencing directly was
done under the auspices of the Institute for Defense Analysis (IDA), beginning in the
early 1960s. The focus of this work was on the possible use of such communications
media as telephone, teletypewriter, and/or television in international relations. Of special
interest was the potential for using teleconferencing in crisis -negotiation situations. This
series of studies, which has only recently been released to the general public, can be
considered as a kind of methodological forerunner of the work which is described in this
article.
The theoretical work done by the IDA is still instructive for research design
involving group communication. Figure 1 shows the key elements identified in these
510
R. Johansen, R. H. Miller, and J. Vallee
studies. Their approach involved simulated crises in laboratory situations and field trials
using different combinations of media. Since another of the purposes of the IDA studies
was to "assemble and review information relevant to teleconferences and teleconference
research and to draw implications therefrom for the long-range teleconference research
program" [6], this seems an excellent starting point in surveying the current situation.
The IDA studies offer findings which are uniquely geared to international crisis
situations, but which can also be generalized to some degree. For instance, the speed of
communication offered by telemedia was thought to be an advantage, but was later found
to have negative effects in those negotiation situations where participants needed time to
think before responding [7]. Also, media which encourage rigid behavior patterns (e.g.,
the teletypewriter, where all communication is via print) were found to increase the
need for parallel channels to provide for informal and symbolic exchanges. "It appears,
then, that historical biases, formal and stilted language, and a disposition to defend
one's position to great lengths are characteristics that are potential drawbacks to
effective communications by written message [8]. However, the IDA simula tion of a
teletype message forwarding system does not allow the extrapolation of these
conclusions to the computerized conferencing systems which permit more than two
people to interact simultaneously.
Current Research Approaches
The social research which is currently being done with personal communications media
can be divided into three, sometimes overlapping, categories: laboratory experiments;
Fig. 1. Typical predictive system for the experimental study of teleconferencing (from
IDA teleconferencing studies [9]).
Computer/Future of Delphi: Group Communication
511
field trials; survey research.
Each of the approaches has certain benefits and weaknesses, some of which we
shall try to point out. Also, such an arbitrary division seems necessary as a first step in
defining possibilities for comparing results and interpreting the research findings which
will become increasingly commo n over the next few years.
Laboratory Experiments: The most classic of these research approaches arises out of
the traditions of experimental psychology. The goal here is the control and
manipulation of certain key elements (independent variables), while monitoring the
resultant effect on other elements (dependent variables). Because of the problems in
monitoring the many variables surrounding a social situation, laboratories are used to
establish a controllable environment. Fromthis point, attempts are made to design the
laboratory in such a way that it replicates (or at least approximates) the "real world."
In the case of communications research, the problems of control have been
magnified. In even the most "simple" instances of interpersonal communication,
multiple complexities are always present. A researcher must attempt to isolate the
effects of a communications medium from the interrelated effects of such things as
group dynamics, personal attitudes, and topical content of the communication. In a
situation such as this there is the constant danger of simplifying the "real world" to
meet the limitations of the laboratory.
Bell Laboratories has produced much work in communications and informa tion
theory [101 and this work has continued, using variations in experimental
methodology. The work at Bell Labs is frequently tied to the development of new
communications technology. There is, however, a renewed interest in the exploration
of basic communications processes -apart from the application of a specific technology.
For instance, ongoing work at Bell Labs is now concentrating on the behavioral
dimensions of two-person, face-to-face communication, with an eventual goal being
the development of a procedure for comparing and evaluating different media of
communication. This work is strengthened by interesting applications of statistical
techniques (particularly multidimensional scaling) to the unique characteristics of
interpersonal communication through electronic media [11].
In 1970 the Communications Studies Group (CSG) was founded in London,
England, with direct support from the Civil Service Department and the Post Office
[12]. CSG has now become a major center of telecommunications research, using a
style begun with a base in laboratory experiments and mathematical modeling. CSG
has also begun exploration of attitudes toward various communications media -a
dimension of communications research which has been largely neglected. The
experiments done by CSG are, of course, considered within the context of actual
problems and planning in the English government.
Several general conclusions from CSG's Final Report (September 1973, Vol. 1)
provide an overview of some current results:
?? Two criteria characterize tasks whose outcome is likely to be affected by
medium of communication: the task must necessitate interaction and must be
such that personal relationships are relevant to the outcome. Thus
communication involving negotiation or interpersonal relations between the
512
R. Johansen, R. H. Miller, and J. Vallee
participants forms areas of sensitivity. Information exchange and problem
solving were two important purposes for which the outcome, in two-person
tasks, was found to be insensitive to variation in the medium of
communication.
?? Attitudes toward media are dependent on the tasks for which they are being
used.
?? A substantial number of business meetings which now occur face-to-face could
be conducted effectively by some kind of group telemedia (usually not merely
the telephone) [13].
As can be seen, the conclusions are quite general at this point, but CSG also has a
growing amount of data on particular experimental situations.
Alphonse Chapanis at the Johns Hopkins University has been doing laboratory
research "aimed at discovering principles of human communication that may be useful
in the design of conversational computers of the future" [14]. Though his facilities are
quite limited at this point, Professor Chapanis has added much toward identification of
dependent variables for evaluating communication patterns in laboratory settings. To
date, his experiments have been exclusively centered on two-person communication
and experimental tasks which have a defined solution. His plans, however, are to move
into group experiments with more open-ended experimental tasks.
Chapanis has done a series of laboratory experiments comparing audio,
handwritten, teletypewritten, and face-to-face communication. The tasks were carefully
selected to be credible "real-world" situations, but the two communicators were always
identified as "seeker" and "source." Thus the experiments actually use informationseeking and information-giving tasks [15]. In this test environment, the results showed
that the oral media (audio and face-to-face) were clearly much faster for solving the
test problems than were handwriting and typewriting. Much more information could be
passed back and forth in the oral modes. (Als o, a general finding has been that level of
typing skill per se has little effect on the generally slower communication time.) These
conclusions suggest that nonverbal communication media offer a more restricted
environment than do the more common oral media. Replication studies which broaden
the group and task components of the Chapanis work seem crucial for the near future,
however, if these results are to be accepted on a more general level.
It should be noted that laboratory experiments involving communication process
have typically concentrated on two-person communication, with clearly defined tasks.
(The limitations of this approach are discussed later in this article.) Thus, time to
solution of the task is often a major criterion. Also, the inherent problems of simulating
the "real world" in a laboratory are especially intense when trying to facilitate a
"natural" communication process in an artificial environment. There is rarely any
continuity in this environment, meaning there is no prior communication or follow-up
to the actual communication situation being evaluated. These factors raise validity
questions about the experimental approach, though the approach certainly has its
appealing aspects (e.g., higher degree of control and ability to isolate key factors).
Computer/Future of Delphi: Group Communication
513
Field Trials: In order to clarify the distinction between laboratory and field
experiments, it seems most appropriate to touch briefly upon the theoretical
characteristics of a quasi experiment:
There are many natural social settings in which the research person can introduce
something like experimental design into his scheduling of data collection procedures
(e.g., the when and to whom of measurement), even though he lacks the full control
over the scheduling of experimental stimuli (the when and to whom of exposure and
the ability to randomize exposures) which make a true experiment possible.
Collectively, such situations can be regarded as quasi-experimental designs [16]."
For our purposes here, field experiments are defined as explorations of actual
"real-world" situations with a minimum of experimental manipulation. In this' sense
they are quasi-experiments, though considerations such as randomized' sampling are
usually not involved. Thus, in general, some of the techniques of the laboratory are
applied under less controlled circumstances.
Such a field experiment in electronically mediated group communications was
performed at Carleton University (Ottawa, Canada) under the auspices of the
Department of Communications, Canada. Jay Weston and Christian Kristen were the
principal investigators in this exploratory attempt at developing "appropriate
methodologies and measures for evaluating the behavioral effects and effectiveness of
broadly defined teleconferencing systems" [17]. The Weston-Kristen effort involved
direct comparison of three communications media (face-to-face, mediated video plus
graphic video, and mediated audio plus graphic video), which were used as a basic part
of the pedagogy in a human communications course. Thus, the experiment was actually
done as part of the students' normal academic program, as they participated in the
group-conferencing sessions. The data were then gathered from self-reporting
questionnaires, analysis of verbatim transcripts, and analysis of split-screen videotapes
of the sessions. Of course, the techniques for performing these analyses are in some
cases rather undeveloped and exploratory in themselves (e.g., content analysis of
transcripts and videotapes). However, our research has revealed very few comparable
efforts at analysis of group communication through alternate media. Thus this effort
should become an important prototype.
Our own work at the Institute for the Future has been moving toward a field
experiments model in the analysis of the social effects of computerized conferencing
[18]. These experiments take a somewhat different tack, since they begin with the task
of developmental work on this new medium of communication-while still striving for
established criteria which can be used to compare computer teleconferencing with other
media. With a developing medium, research problems are further complicated, since
the results will almost certainly vary as the characteristics of the medium evolve. (In
fact, the results of tests will actually influence this evolution.) Also, one of our
supporting grants comes from the National Science Foundation to explore the potential
for using computerized conferencing to improve interaction among experts. Not surprisingly, it is difficult to explore "expert interaction" in laboratories, since one of the
major characteristics of "expertness" is the availability of personal resources (files,
514
R. Johansen, R. H. Miller, and J. Vallee
Fig. 2. Categorization of measurement techniques for analysis of computer conferences
(Institute for the Future).
Computer/Future of Delphi: Group Communication
515
library, etc.) - even assuming that we could convince experts to come into a laboratory
setting. Thus the experimental problems surrounding the development of new media
add greatly to the already pressing problems of group research using electronic media.
The research design for the above project began with the gathering of folklore
about the new medium, since we really didn't even know what questions to ask without
biasing reactions. Then the approach gradually became ,more systematic, working
toward the field experiments model. Thus our procedure moved from gathering
anecdotal information to field experiments with defined user groups and systematic
collection of results using multiple measures of effectiveness primarily user reaction
and attitudes, content analysis where appropriate, and analysis of communications
patterns [19]. (See Fig. 2.)
Another example of the field experiment approach is that applied by the British
Columbia Telephone Company in their initial tests of the Confravision system for
video conferencing. This system, given the same name as that used by the British Post
Office, provides live television links between two conference rooms located in
Vancouver and Victoria. Under the direction of Anders Skoe, a process for initial field
experiments was developed and implemented before the system was to be made
generally available. These field experiments involved selected types of users who were
representative of those expected to use the system. The evaluation procedure employed
the following techniques: preexperiment interview; live observation during the test
with report by independent consultant; posttrial questionnaire (based on the
"Teleconference User Opinion Questionnaire" developed by Communication Studies
Group); questionnaire/interview some weeks after test session [20].
In this way the British Columbia Telephone Company group hoped to begin
assessing the behavioral impact of the medium on the people who would be using it.
Also, this behavioral goal was linked to initial technical tests of the 'system and longerrange socioeconomic forecasts.
An important characteristic of field trials is the ability to run tests over a longer
period of time. Thus the test itself becomes more credible, since it is conducted with a
sense of continuity and integration with everyday experiences. One example of such a
field trial is that being conducted by the New Rural Society Project in Stamford,
Connecticut. In this trial, two banks located in Stamford and New Haven are connected
via an audio conferencing system, with a studio at each location. The time period for
the trial is six months, and a battery of questionnaires was developed to assess both
expectations of the system and reactions at various points in time.
The field trial approach, then, attempts to apply experimental research
procedures to the degree which they can be effectively used in the "real world."
Though the controls are limited, the goals are to gather a maximum amount of
systematic data in actual group communication situations. These field trials can vary
from quasi-experiments, which have a higher degree of control over variables, to the
more open-ended kinds of field trials.
Survey Research: The basic tools of survey research are the old reliables (?):
questionnaires and interviews. In communications research of this sort, however,
unique problems are added to the routine dilemmas of the survey researcher. For
516
R. Johansen, R. H. Miller, and J. Vallee
instance, subjects might be asked to evaluate their needs for media which they have not
yet experienced. Still,, the techniques of survey research remain the most basic of the
social sciences, and can be used creatively to gather information on both reactions to
existing media and speculations about future needs.
A good example of survey techniques in telecommunications research is found in
the study by Dean Havron and Mike Averill, which had as its goal "to develop a plan
and instrument for a survey of needs of Canadian Government managers fo r
teleconference facilities and equipment" [21]. This goal was pursued by designing a
questionnaire which was administered to potential teleconference users, asking them
a series of questions related to their present conferencing style and the possibilities
for employing teleconferencing. The specific strategy of the questionnaire was to ask
the respondents about a meeting they attended recently which required travel on the
part of some group members. From this base, respondents were then asked to project
what (if any) teleconferencing facilities could be used to conduct a future meeting of
the same sort.
A similar project was directed by James Kollen of Bell Canada [22]. Kollen
focused on existing travel patterns between Montreal, Quebec City, Ottawa, and
Toronto. A questionnaire was administered to businessmen traveling between these
cities, with the goal of determining why respondents felt they needed to travel (rather
than use alternate communications media) to achieve their objectives. In this case,
the overall purpose of the project was to explore the possibilities for substituting
electronic communication for travel in interurban exchanges.
Dean Havron, with Anna E. Casey-Stahmer, was also involved in a project
which used survey techniques to assess existing telecommunications systems [23]. In
an effort to establish a general approach to research in teleconferencing, they
developed a grid to classify teleconferencing system dimensions and characteristics
through information gathered in interviews with actual users. Also involved were a
series of interviews with users of four existing tele conferencing systems.
The techniques of survey research are certainly relevant to the social evaluation
of communications media, and are frequently employed-even in mo re controlled
situations such as those mentioned earlier in this article. Figure 3 summarizes the
various research approaches described earlier.
Isolating Key Elements in Mediated Group Communication
The basic research approaches outlined here adopt different methodologies for
approaching common research problems. In efforts to assess any communications
medium, however, some comparison with other media (usually with face-to -face) is
typically assumed and is certainly of central importance. Yet in order to make such
comparisons, a duplicate series of techniques is less important than commonality in
the research philosophy which prefaces the choice of methodology. In fact, any of
the general approaches outlined in this article could be justified as valid tactics, and
methods which cross these suggested categories may also be appropriate. The
problem in comparison comes in the adoption of a general taxonomy which can be
employed across media in various group communications situations.
Computer/Future of Delphi: Group Communication
517
NOTE: The classifications shown, here represent our perception of the
dominant research styles of various groups which have been active
at some time.
No doubt there will be some overlap in the categories.
Fig. 3 . Summary of approaches.
518
R. Johansen, R. H. Miller, and J. Vallee
The taxonomy which we are suggesting would simply frame the most
fundamental questions which must be asked in order to assess a particular
communication situation involving any group. These questions must apply across
media (including face-to -face) and they must be flexible enough to encourage
development of a broad range of research techniques.
The existing taxonomies of group process, however, tend to be oriented toward
dyadic communication (only two persons), and extrapolation from dyadic patterns to
group patterns of communication seems to be questionable. The principles simply
cannot be assumed to be transferable, though they certainly should not be ignored.
Our own examination of the literature reveals that perhaps the most useful
taxonomy was developed in the context of the recently declassified tele conferencing
studies by the Institute for Defense Analyses [24] which were mentioned earlier.
These studies examined the possibilities for using various kinds of teleconferencing
systems in international crisis communications of the sort typical of NATO. Thus
their efforts focused on the careful evaluation and comparison of various
teleconferencing systems in relation to the peculiar characteristics of crisis
negotiations (e.g., translation problems, time constraints, etc.). The key variables
were divided into independent variables (e.g., tele conference arrangements,
dimensions of crises, and social determinants), intervening variables (e.g., interaction
process), and criterion variables (e.g., group satisfaction and group outcome) [25].
Our own attempt to construct a taxonomy of this sort is similar to the IDA
effort, but attempts to be more precise and does not assume the crisis orienta tion.
Also, our taxonomy of elements does not attempt to incorporate the dynamic aspects
of the communication. Through this does, of course, need to be analyzed, the
taxonomy merely attempts to isolate the elements in a communication situation
before the interpersonal personal process begins.
The taxonomy (shown in Fig. 4) is arranged to suggest a varied weighting
among five key factors. None of the factors will be completely discrete. For instance,
if members of a given group have a very high need to communicate, they are more
likely to make appropriate efforts to gain access to the chosen medium-even if it is
difficult to use or unfamiliar to them. Conversely, however, familiarity with a
particulat media is likely to be a very important factor in choice of that media, unless
some other factor becomes more important.
The problem in constructing a taxonomy of this sort is making it flexible
enough to include all options, while still keeping its utility in terms of decision
making. The taxonomy should provide basic advice about such things as choice of
media when a series of options are available. For instance, Rudy Bretz has
constructed a taxonomy of communications media which divides them into eight
classes according to the coding process which is being employed for sending
messages (e.g., audio/motion/visual, audio only, print only, etc.). Having established
this taxonomy, he then traces the necessary decision points in making a choice of the
simplest medium available to fill a specific instructional need [26]. When viewed
within the context of the general taxonomy of group communication suggested in this
article, the decision is based within the sections labeled Medium of Communication
Computer/Future of Delphi: Group Communication
519
Fig. 4 . Identification of elements in a group communication situation (before
process actually begins).
Fig. 4 . Identification of elements in a group communication situation
(before process actually begins).
520
R. Johansen, R. H. Miller, and J. Vallee
and Task. The dimensions of Individual, Group/Aggregate, and Rules for Conduct are
thus not included in the Bretz taxonomy.
The question, of course, is determining the most fundamental data needed to
make the best possible decision. The Bretz taxonomy moves away from generality in
order to develop specific sequence of questions and typologies. This process seems
inevitable, though the actual choice of fundamental factors remains debatable.
Another approach to selection of media for instruction was developed by C.
Edward Cavert [27]. This is a very usable technique which operates from a matrix
showing available media on one axis, with a modification of Bloom's taxonomy of
learning on the other axis. (The latter is a generally accepted technique for classifying
types of learning, from note memorization to ability to apply knowledge in new
contexts.) Cavert then suggests that each available medium be matched to the level of
learning with which it seems most appropriate, and that the overall media strategy of a
school involves a distribution of available media across types of learning. Thus he has
developed a simple technique for assessing the general choices of media for various
needs within some overall design.
Another basic taxonomy, which differs somewhat in its approach to media, is that
developed by Fred Lakin [28]. This taxonomy functions like a kind of coding sheet,
which includes a general format for classification and possible exceptions. Though this
approach is not yet developed for choice among media, it represents a good general
format which is quite practical in doing initial classifications of media.
As was mentioned previously, the existing taxonomies of group process and most
experimental data concerning communication process are based on two-person
communication. Even though a number of experiments and field trials have been
categorized as "group" communication, most of these teleconference systems have
dealt with the interconnection of two face-to-face groups (i.e., where an individual is in
face-to-face contact with his own group and in contact via electronic media with a
single distant group) [29].
With the exception of experiments involving conference telephone [30], little
consideration has been given to the theoretical (i.e., in terms of taxonomy) or
behavioral aspects of group interaction solely by means of a teleconferencing medium.
This situation is analogous to the three-body or n-body problem in physics, in which
description of two-body interactions does not reveal anything about three-body or nbody interactions.
Similarly, little research besides the communication network studies by persons
such as Bavelas or Smith and Leavitt has dealt with any of the behavioral or theoretical
aspects of the n-body problem of communications.
Future research with computer-assisted media such as FORUM demands
investigation of this n-body problem, since the typical scenario includes several users,
each connected to others by means of a terminal. The peculiarities of computer
conferencing do not allow a facile transfer of theoretical or behavioral findings to other
media due to differences in (a) time and synchronicity of interaction, and (b) the code
(i.e., typewritten text) used for communication. Since it seems evident that the
interaction of three or more individuals by means of different electronic media is
affected by the limits and constraints of technology and human engineering, the
Computer/Future of Delphi: Group Communication
521
problems of n-body communications should be a high priority subject for future
investigation.
Another technological consideration in studying group interaction via electronic
media is "medium memory"; that is, the capability of a medium to produce a record of
an interaction which is available to the user. Information theory and communication
theory have often dealt with "channels with memory," but this has been only on a
technological/engineering basis. The various behavioral aspects of "medium memory"
(e.g., its presence or absence, its availability to users, the extent of the record, etc.) may
produce significant effects in the use and perception of teleconferencing media and
thus must be examined in future research.
The question of taxonomies is frustrating, while still remaining basic to any hope
for exchange of evaluations of various media. In pursuit of this goal, a small workshop
was held at the Institute for the Future in February of 1974. One of its goals was to
agree upon a general taxonomy for the assessment of group communication, with an
emphasis on electronic media [31]. The workshop group was intentionally kept very
small in order to promote maxi mum interaction, and thus was not inclusive of all
persons doing key research in this field. However, it was hoped that this initia l face-toface workshop would build the basis for broader exchange of research approaches and
results in a more continuous manner (probably using some telecommunications media).
An earlier draft of the elements shown in Fig. 2 was used as a basis for the
workshop discussion of taxonomies. Our interpretation of the reactions expressed at
that time is that the chart of elements identifies the basic issues -but doesn't deal with
the vital factor of communication process. Though this was not the purpose of the
chart, it is intended that the elements be used within a larger analysis of communication
over time. It seemed that all the workshop participants were agreed that a sense of
movement-including some analysis of before and after the actual meeting-was perhaps
the most crucial need in current media research. Research methods have tended to be
too static, and must somehow develop an ability to consider dynamic factors in group
process.
Such a realization prompted several specific suggestions for the near future of
research. One key suggestion was that our methodology must draw more heavily from
fields such as anthropology, since the kind of input which is needed moves beyond the
realm of traditional social-psychological research. It was also decided that a core set of
questions (not a "universal questionnaire") should be developed and made available to
anyone concerned with evaluating group communication situations. These questions
could have an important effect on the ability to compare research results across media.
Other methods of continued exchange among researchers were suggested, and specific
conferences using the FORUM teleconferencing system have now begun.
Conclusion
The approaches outlined in this article all attempt various forms of systematic appraisal
of media, using a range of formality. In the field of telecommunications, systematic
evaluation of social effects has not been a broadly accepted practice. Rather, there has
been a strong tendency toward initial expenditures on technology, with social impact
522
R. Johansen, R. H. Miller, and J. Vallee
analysts used only in later stages of imple mentation-and then only sparingly. The
almost nonexistent literature on the sociology of the telephone is a prime example of
what has now become a norm in the field of telecommunications. Thus, the approaches
presented in this article represent a variety of techniques which-even when considered
as a group-may amount to an (as yet) insignificant input to the future of telecommunications research. The range of activities described here, though, suggests that
interest in social implications of telecommunications media is both growing and
increasing in sophistication. As these techniques become more effective-and assuming
that communications channels within the various research communities continue to
develop-there is a hope for offering more intelligent answers to questions of media
usage. The mysteries of human communication will remain dominant, but the ability to
choose a comple mentary medium of communication seems likely to improve as the
inherent "messages" of various media become known and more effectively directed.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Anna E. Casey-Stahmer and M. Dean Havron, Planning Research in Teleconference
Systems, Human Sciences Research Institute (November 1973).
Joseph E. McGrath and Irwin Altman, Small Group Research (New York: Holt, Rinehart &
Winston, 1966).
Alex Bavelas, Teleconferencing: Background Information, Research Paper P-106, Institute
for Defense Analyses (1963), p. 4.
Ibid., p. 12.
Alex Reid, New Directions in Telecommunications Research, a report prepared for the
Sloan Commission on Cable Communications (June 1971).
Gerald Bailey, Peter Nordlie, and Frank Sistrunk, Teleconferencing. Literature Review,
Field Studies, and Working Papers, Research Paper P-113, institute for Defense Analyses
(October 1963), p. 2.
Ibid., p. 5.
W. Richard Kite and Paul C. Vity, Teleconferencing: Effects of Communication Medium,
Network and Distribution of Resources, Study S-233, Institute for Defense Analyses.
Bailey, Nordlie, and Sistrunk, op. cit., pp. 1-19.
E.g., Claude E. Shannon and Warren Weaver, The Mathematical Theory of
Communications (Urbane, Illinois: University of Illinois Press, 1949).
See J. Douglas Carroll and Myron Wish, "Multidimensional Scaling: Models, Methods,
and Relations to Delphi," Chapter VI, C, of this volume.
Communications Studies Group is a part of the joint Unit for Planning Research of
University College, London, and the London School of Economics.
Communications Studies Group; Final Report, September 1973, Vol. 1, a report
submitted to Management Services Division of the Civil Service Department and the
Long Range Studies Division of the Post Office, p. 2.
Alphonse Chapanis, "The Communication of Factual Information Through Various
Channels," Information Storage and Retrieval 9, p. 215.
The classification of tasks is an essential problem in the research of media for group
communication. Though no standard typology has yet been adopted by the entire field,
the Communications Studies Group has developed one framework called Description
and Classification of Meetings (DACOM). Eleven types of meeting tasks are
identified, including tasks such as information exchange, .problem-solving, conflict of
Computer/Future of Delphi: Group Communication
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
523
interest, etc. Bell Labs at Murray Hill, New Jersey, has also done some work in this
area and there is a standard typology oil tasks which was developed at Harvard. Some
agreement on typologies of task will be an important factor in data exchange among
researchers in the future.
Donald T. Campbell and Julian C. Stanley, Experimental and Quasi-Experimental
Designs for Research (Chicago: Rand McNally, 1963).
J. R. Weston, Teleconferencing and Social Negentropy, presented to International
Communication Association, Montreal, Quebec (April 1973), p. 8. See also, J. R.
Weston and C. Kristen, Teleconferencing: A Comparison of Attitudes, Uncertainty and
Interpersonal Atmospheres in Mediated and Face-to-Face Group Interaction, The Social
Policy and Programs Branch, Department of C o m munications, Ottawa, Canada
(December 1973).
Computer teleconferencing is a medium of communication where people "talk" through
typewriter keyboards connected to a central computer-often through a large computer
network. The participants communicate in standard English, with the computer acting
as a connecting device and storing the information in the form specified for the
conference. The group may be large or small, and may be present simultaneously or
whenever each individual prefers. This medium was first conceived by Dr. Murray
Turoff, now at New Jersey Institute of Technology.
See Group Communication through Computers, Vols. 1 and 2, Institute for the Future
(Summer, 1974). Volume 2 describes the current techniques being used in the
analysis of FORUM conferences.
Personal correspondence from Anders Skoe, Sociological Analyst, British Columbia
Telephone Company, September 25, 1973.
M. Dean Havron and Mike Averill, "Questionnaire and Plan for Survey of
Teleconference Needs among Government Managers," prepared for the SocioEconomic Branch, Department of Communications, Canada, Contract OGR2-0303
(November 30, 1972), p. 1.
See James H. Kollen, "Transportation-Communication Substitutability: A Research
Proposal," Bell Canada (February 1973).
Casey-Stahmer and Havron, op. cit.
Bailey, Nordlie, and Sistrunk, op. cit.
Ibid., p. 14.
Rudy Bretz, The Selection of Appropriate Communication Media for Instruction: A Guide
for Designers of Air Force Technical Training Programs, Report R-601-PR, the Rand
Corporation (February 1971), pp. 30ff.
See C. Edward Cavert, Procedures for the Design of Mediated Instruction, State
University m . Nebraska Project (1972).
See Fred Latin, Media for a Working Group Display (Fred Latin, 218 Waverley St., #C,
Palo Alto, California 94301).
For instance, the Canadian Department of Communication Audio-Graphic System, Bell
Canada Conference TV, and the First National City Bank Audio-Video Teleconferencing
System, all described in Casey-Stahmer and Havron, op. cit See also, Communications
Studies Group, The RMT Tekconference System, P/72024/RD (1972).
See Communications Studies Group, Progress in Current Experiment, W/71132/CH (1971);
Communications Studies Group, Bargaining at Bell Laboratories, E/71270/CH (1971); and
Casey-Stahmer and Havron, op. cit.
The participants in this workshop were:
Garth Jowett, Department of Communications,
Canada; Lee McMahon, Bell Laboratories; Martin Elton, Communications Studies Group;
Jay Weston, Carleton University; Christian Kristen, University of Montreal; Robert
524
R. Johansen, R. H. Miller, and J. Vallee
Johansen, Richard Miller, and Jacques Vallee, Institute for the Future; George Jull and
James Craig, Communications Research Center, Canada; Mike Averill, National Film
Board of Canada; Tony Niskanen, Arthur D. Little, Inc.; and Dean Havron, Human
Sciences Research, Inc.